POSITIVISM AND LEGALITY: HART’S EQUIVOCAL RESPONSE TO FULLER · TO FULLER JEREMY WALDRON* Lon...

\\server05\productn\N\NYU\83-4\NYU407.txt unknown Seq: 1 25-SEP-08 12:34

POSITIVISM AND LEGALITY:HART’S EQUIVOCAL RESPONSE

TO FULLER

JEREMY WALDRON*

Lon Fuller, in his response to H.L.A. Hart’s 1958 Holmes Lecture and elsewhere,argued that principles of legality—formal principles requiring, for example, thatlaws be clear, general, and prospective—constitute the “internal morality of law.”This Article contends that Hart never offered a clear response. Fuller’s claim sup-poses that observance of the principles of legality is both fundamental to law andinherently moral. In different writings, Hart seems variously to affirm and to denythat legality is a necessary criterion for the existence of law. Likewise, he sometimessuggests and elsewhere scorns the idea that legality has moral significance. ThisArticle proposes that Hart’s apparent inconsistency might actually reflect the com-plexity of the terms. Some degree of legality might be a prerequisite of law, whilesome failures of legality might not condemn it. Principles of legality might havecontingent rather than inherent moral value, might have moral value that is sever-able from their legal value, or might have both positive and negative moral effect.The Article argues, furthermore, that even the conclusion Hart strains to avoid—that legality inevitably links morality and law—is compatible with Hart’s positivismand opens a promising field for positivist jurisprudence.

INTRODUCTION

One of the most telling observations that Lon Fuller made in his1958 response to H.L.A. Hart’s Holmes Lecture 1 concerned Hart’sapparently blinkered view of the evils of rule by Hitler and the Naziparty in Germany from 1933 to 1945. Fuller said this: “Throughouthis discussion Professor Hart seems to assume that the only differencebetween Nazi law and, say, English law is that the Nazis used theirlaws to achieve ends that are odious to an Englishman.”2 Of coursethere was no disagreement between Hart and Fuller about the odious-ness of the ends that the Nazis pursued and also the wicked means

* Copyright 2008 by Jeremy Waldron, University Professor, New York UniversitySchool of Law. This Article was presented at the Symposium on the Hart-Fuller Debate atFifty, held at the New York University School of Law on February 1–2, 2008. I am gratefulto the participants, particularly Jules Coleman, Les Green, and Nicola Lacey, for theirsuggestions. Thanks also to Ben Kingsley and Sandy Mayson for their most helpful com-ments and suggestions in the editing phase. An earlier and much shorter version of thisArticle was presented at a conference on “The Legacy of H.L.A. Hart” at Cambridge Uni-versity in August 2007.

1 H.L.A. Hart, Positivism and the Separation of Law and Morals, 71 HARV. L. REV.593, 595–96 (1958).

2 Lon L. Fuller, Positivism and Fidelity to Law—A Reply to Professor Hart, 71 HARV.L. REV. 630, 650 (1958).

1135


1136 NEW YORK UNIVERSITY LAW REVIEW [Vol. 83:1135

through which they pursued them: national aggrandizement, racialsupremacy, aggressive war, genocide, and the use of murder, terror,torture, and reprisals as routine modes of political control. Everydecent person recoils from the memory of these horrors.

But for Fuller there were also aspects of Nazi misrule thatdeserved the special attention of legal philosophers. One was the factthat along with the substance of the murderous Nazi tyranny camesustained violations of principles of legality—for example, violationsof principles requiring prospectivity, publication of laws, elementarylegal process, and legal restraints upon agencies of the state. Fullerthought that this particular aspect of the Nazi tyranny, which (in thefirst instance) concerned forms and procedures rather than ends andpurposes, ought to be of special concern for jurisprudence because, hesaid, it might affect our willingness to describe Nazi rule as rule bylaw:

When a system calling itself law is predicated upon a general disre-gard by judges of the terms of the laws they purport to enforce,when this system habitually cures its legal irregularities, even thegrossest, by retroactive statutes, when it has only to resort to foraysof terror in the streets, which no one dares challenge, in order toescape even those scant restraints imposed by the pretence oflegality—when all these things have become true of a dictatorship,it is not hard for me, at least, to deny to it the name of law.3

In addition, Fuller thought it worth exploring the possibility thatthese formal and procedural violations might have a substantive moralsignificance—a significance that encouraged him to describe what Ihave called “principles of legality” as principles of “the morality oflaw itself,”4 law’s “own implicit morality,”5 “the internal morality oflaw,”6 and “the inner morality of law.”7 Fuller reminded his readersthat governance in Germany in those years was not uniformly afflictedby these formal and procedural defects; private law was not affectedin the same way or to the same extent. It was in the area of race lawsand laws governing the operation of the political system (such as itwas) that there was this tendency toward secrecy, retroactivity, andthe repudiation of legal restraints:

It was in those areas where the ends of law were most odious byordinary standards of decency that the morality of law itself wasmost flagrantly disregarded. In other words, where one would have

3 Id. at 660.4 Id. at 644.5 Id. at 645.6 Id.7 Id. at 650.


October 2008] HART’S EQUIVOCAL RESPONSE 1137

been most tempted to say, “This is so evil it cannot be a law,” onecould usually have said instead, “This thing is the product of asystem so oblivious to the morality of law that it is not entitled to becalled a law.” I think there is something more than accident here,for the overlapping suggests that legal morality cannot live when itis severed from a striving toward justice and decency.8

And so Fuller thought it worth entertaining the hypothesis that asystem that abides by principles of legality is less likely to be com-mitted to the sorts of odious ends that the Nazis pursued—or, if it iscommitted to those ends, is less likely to be able to pursue them asthoroughly as the Nazis did.

Fuller’s reflections on these matters suggest a twofold agenda forjurisprudence. It might be worth asking, first, “What exactly is therelation between the principles of legality and the categories law andlegal system that we use to characterize systems of rule?” And itmight be worth asking, second, “What exactly is the relation betweenprinciples of legality and norms like justice, rights, and the advance-ment of the common good that we use to evaluate systems of rule?”Fuller’s 1958 response to Hart’s Holmes Lecture argued that thesequestions were worth asking for the sake of a subtler jurisprudentialdissection of the Nazi horror than Hart seemed willing to undertake.But they might also be worth asking about legal systems and systemsof rule in general. Fuller thought that asking and answering thesequestions promised to enrich the categories used in legal and politicalphilosophy. But legal positivists have shied away from this suggestionout of fear that it might open the boundaries between legal and polit-ical philosophy and complicate our sense of the separability of law andmorality. Indeed, this seems to have been the response of H.L.A.Hart. For although Hart acknowledged that what I have called theprinciples of legality formed an interesting and distinctive subset ofthe principles deployed in legal and political philosophy,9 he neveropenly and unequivocally addressed the two questions I have identi-fied above for fear that the answers might undermine one of the dis-tinctive pillars of his own jurisprudence.

I think Hart was inclined to see a preoccupation with legality andthe rule of law as a source of confusion in jurisprudence; often onegets the impression that Hart thought that if anyone offered to talkabout it, the responsible thing to do was to say something palliativeand then shut down the discussion as quickly and firmly as possible.Principles of legality, Hart implied, may be among the principles we

8 Id. at 661.9 See infra Part II.



should use for the evaluation of law, but their study is not part of thephilosophical discipline that tries to tell us what law essentially is.

I do not mean that Hart was hostile to the rule of law as a polit-ical ideal. Neil MacCormick once became very indignant about a sug-gestion made by Fuller to the effect that Hart embraced a“managerialist” approach to law.10 Professor MacCormick said this inresponse:

Nobody who gave five minutes’ cursory thought to Hart’s variousbut largely self-consistent reflections about the moral relevance ofpositive law . . . . could suppose that he is any less an enthusiast thanL.L. Fuller for promoting the vision of a society in which freelycommunicating individuals willingly collaborate in their commonsocial enterprises and freely grant each other friendly tolerance intheir more particularistic or individual activities, and in which theresort by officials to means of mere coercion is minimized. Nobodycould deny either the reality of his concern for justice or the firm-ness of his contentions that a precondition of justice as definedwithin his critical morality is the existence of a well working legalsystem and that a consequence of a just legal system’s existence isthe establishment of a network of mutual moral obligations ofrespect for law among the citizens within that jurisdiction.11

It is not Hart’s personal enthusiasm for the rule of law, however,that one misses in his jurisprudence. What one seeks is an elucidationof it and some basis for concluding that legality is a topic worthy ofjurisprudential analysis. And what one finds is mostly equivocation.

In this Article, we will attempt to pin down what exactly Hartthought about legality. In Part I, we will sharpen our understandingof the two questions that Lon Fuller’s comments on legality suggestfor Hart’s jurisprudence. In Part II, we will see that Hart certainlyacknowledged the existence of principles that were generally called“principles of legality,” though it is not clear whether he himself wascomfortable calling them that. In Part III, we shall look in detail,piece by piece, at what Hart said about these principles (whatever onecalls them). In Part IV, we shall examine more systematically whatappear to be inconsistencies in Hart’s treatment of legality. And inPart V, we shall consider whether those inconsistencies are real orwhether the semblance of inconsistency conceals a deeper coherencein his position.

As we shall see, Hart’s comments about Fuller were rather dis-missive. But apart from his dismissive tone, there is considerable

10 NEIL MACCORMICK, H.L.A. HART 157–58 (1981) (citing LON L. FULLER, THE

MORALITY OF LAW 207–13 (rev. ed. 1969)).11 Id.



uncertainty about what Hart wanted to say in response to the ques-tions that Fuller’s discussion raises and what he would have said hadhe been willing to confront these questions explicitly.

ITWO QUESTIONS ABOUT LEGALITY

I said above that Fuller’s comments raised two important ques-tions. The first is about the relation between legality and law:

Is there an important relation between the principles of legality andthe concepts of law and legal system?

The second question is about the relation between legality andmorality:

Is there an important relation between the principles of legality andmorality?

The questions are important both individually and as a pair. Individu-ally, they help us to understand the concept of legality (and the princi-ples of legality). They are important as a pair because an affirmativeanswer to both might suggest, by transitivity, a connection betweenlaw and morality that would be of major importance (or major con-cern) from the point of view of modern analytical jurisprudence.

But the questions need to be sharpened a little. First, we shouldclarify their domain. They are questions about organized systems ofrule—that is, about social or political systems in which human conductis governed in one way or another. It is important to understand thata system of rule may or may not be a system of law (or rule by law),and that the concepts law and legal system pick out some systems ofrule and distinguish them from others. I do not mean this to beg anyquestions. All participants in the jurisprudential debate recognizethat some human societies organize themselves (or are organized,dominated, or exploited) on some basis other than law. In The Con-cept of Law, Hart talked, for example, about prelegal societies.12 Ofcourse, jurists may differ radically on the parameters of the class ofnonlegal systems of rule. Those who give an affirmative answer to thefirst question—Is there an important relation between the principlesof legality and the definition of a legal system?—may believe that it isa much larger class than those who give a negative answer are willingto acknowledge. For those who give an affirmative answer to the firstquestion, the class of nonlegal systems of rule includes all those sys-tems that call themselves “legal systems” but suffer from significantand pervasive failures of legality.

12 H.L.A. HART, THE CONCEPT OF LAW 91–94 (2d ed. 1994).



A second point is that the term “important relation” is not welldefined. Under its broadest definition, the answer to the first questionis an easy “yes.” Of course there is some logical connection betweenprinciples of legality and the concept of law. Laws are what principlesof legality are designed to evaluate; perhaps principles of legality are(as John Finnis argues) principles for keeping legal systems in goodshape;13 or principles of legality may be designed (as Joseph Razseems to think) to remedy or mitigate evils that only law makes pos-sible.14 But the particular connection between law and legality thatinterests me in this Article is the possibility that Fuller raised: Asystem of rule might depart so far from the principles of legality as toundermine its claim to be called either a system of law or a legalsystem.15 Fuller is suggesting that the principles of legality might berelated criterially to the concepts law and legal system; in other words,they may be among the necessary criteria for the proper application ofthese concepts to a system of rule. So the version of the first questionthat I want to consider is this:

(1) Is observance of the principles of legality among the necessarycriteria for the application of the concepts law and legal system to asystem of rule?

Question (1) is still quite loose inasmuch as it leaves open boththe extent to which the principles of legality must be observed and theappropriate place of this criterion among all the other criteria for theapplication of the concepts law and legal system. Obviously, obser-vance of the principles of legality is not a sufficient condition for theapplication of these concepts (except to the extent that the principlesalready embody reference to other criteria by mentioning rules,courts, etc.). Fuller considered the possibility that the observance oflegality—or as he called it, the “internal morality of law”16—is a nec-essary condition for a system of rule to be regarded as a legal system.But Fuller also acknowledged that observance of legality is a matter ofdegree for several of the principles it includes. Also, legality, on hisaccount, has several dimensions; the various requirements of legalityare not utterly independent of one another but can vary indepen-dently up to a point. In his later formulations, Fuller seemed to sug-gest that nothing less than a comprehensive failure in at least one of

13 JOHN FINNIS, NATURAL LAW AND NATURAL RIGHTS 270 (1980) (“The name com-monly given to the state of affairs in which a legal system is legally in good shape is ‘theRule of Law’. . . .”).

14 JOSEPH RAZ, The Rule of Law and Its Virtue, in THE AUTHORITY OF LAW: ESSAYS

ON LAW AND MORALITY 210, 224 (1979).15 Fuller, supra note 2, at 660.16 Id. at 645.



these dimensions should lead us to withhold the term law from theresulting system of rule.17 We might adopt a tighter or a looser viewthan this. Also, we should pay attention not just to the extent of thefailure in the case of a particular norm or prescription (leading us per-haps to withhold the term law from that norm or prescription inextreme cases) but also to the percentage of the norms or prescrip-tions within the system that were affected in this way.

In the discussion that follows, however, I will not try to pin thingsdown any more than this. We have already sharpened the first ques-tion enough to make it a matter of open controversy between, say,modern positivists and their opponents. We can imagine somemodern positivists—Hart may well be among them—saying thatwhether something is a legal system is one thing, and the extent towhich it satisfies the criteria of legality is another. Such a positionrequires a negative answer to (1), whether the question is aboutFuller’s particular view about the criterial connection or about asomewhat looser view than his. Beyond these observations, there mayalso be some additional indeterminacy in what individual principles oflegality actually mean; I will deal with those issues as they arise in oursurvey and evaluation of Hart’s response to question (1).18

Having noted these considerations, we can imagine two positionsthat a legal philosopher might adopt in response to question (1). Thefirst is a position close to Fuller’s:

(1a) Observance of the principles of legality is among the necessarycriteria for the application of the concepts law and legal system.There comes a point in a system’s failure to observe the principlesof legality where we would have to say that it does not really countas a legal system or a system of law at all.

And the second is the position that some positivists might adopt:(1b) Observance of the principles of legality is not among the neces-sary criteria for the application of the concepts law and legal system.Whether a system of rule counts as a legal system is determinedindependently of whether or to what extent that system observesthe principles of legality.

I shall argue that, in different places, Hart adopts both of thesepositions.

We now turn to the second question: Is there an important rela-tion between the principles of legality and morality? This needs lessin the way of fine-tuning. The bland reference to “morality” requires

17 LON L. FULLER, THE MORALITY OF LAW 39 (rev. ed. 1969) (“A total failure in anyone of these eight directions does not simply result in a bad system of law; it results insomething that is not properly called a legal system at all . . . .”).

18 See infra Parts IV–V.



a little more specification, because we do not want Lon Fulleranswering “yes” to the second question simply on account of his stipu-lative use of the phrase “internal morality of law” to describe the prin-ciples of legality.19 “Morality” in the second question refers to moralvalues like justice, rights, liberty, dignity, and advancement of thecommon good—values that we use in the substantive evaluation ofsystems of rule. If Fuller (or, for that matter, Hart) wants to argue forthe moral significance of legality, he must show a connection betweenthe principles of legality that characterize the operation of a givenlegal system and the substantive moral quality of that legal system. Inother words, the second question asks whether observance of the prin-ciples of legality makes a positive difference to the moral quality ofrule in a given system—e.g., makes it more just, more respectful ofrights, or better attuned to the common good. So we might reformu-late the question as follows:

(2) Does observance of the principles of legality make an affirma-tive difference to the moral quality of a system of rule?

As with question (1), we might be tempted to try to pin down theextent of the observance of the principles of legality that is supposedto trigger a difference in moral quality. It could be that very smallvariations make little or no difference, whereas gross variations arequite significant. I prefer to leave this issue open.

Fuller is famous for having offered a very strong affirmativeanswer to question (2). He suggested that systems of rule that observethe principles of legality are much less likely to be wicked, unjust, ortyrannical. Maintaining that “coherence and goodness have moreaffinity than coherence and evil,”20 that “when men are compelled toexplain and justify their decisions, the effect will generally be to pullthose decisions toward goodness,”21 and that “even in the most per-verted regimes there is a certain hesitancy about writing cruelties,intolerances, and inhumanities into law,”22 Fuller believed in a signifi-cant correlation between legality and justice. This is a very strongposition. But those of us who are less confident in these premises orin Fuller’s conclusion may still give a moderately affirmative answer toquestion (2):

(2a) Observance of the principles of legality tends to make a posi-tive moral difference to a system of rule.

19 See FULLER, supra note 17, at 4 (first appearance of phrase).20 Fuller, supra note 2, at 636.21 Id.22 Id. at 637.



The basis for this moderate affirmative response might be that respectfor the principles of legality is a way of respecting human dignity or ofmaking important concessions to liberty—that is, to people’s ability tocontrol important aspects of their own lives in a meaningful way.

A negative response to question (2) might be motivated either bya denial that the principles of legality make any moral difference or bya belief that the moral difference may in some cases be negative, sothat one cannot count on legality’s making things better. The negativeresponse might be as follows:

(2b) No reliable generalization can be made about the moral differ-ence that observance of the principles of legality makes to a systemof rule.

In addition, there might be some who think that observance ofthe principles of legality tends, reliably and systematically, to makematters worse, morally speaking. This more extreme position mayhave been the position of leftist critics of the rule of law in the past.Thus Morton Horwitz once remarked:

I do not see how a Man of the Left can describe the rule of law as“an unqualified human good”! It undoubtedly restrains power, butit also prevents power’s benevolent exercise. It creates formalequality—a not inconsiderable virtue—but it promotes substantiveinequality by creating a consciousness that radically separates lawfrom politics, means from ends, processes from outcomes. By pro-moting procedural justice it enables the shrewd, the calculating, andthe wealthy to manipulate its forms to their own advantage. And itratifies and legitimates an adversarial, competitive, and atomisticconception of human relations.23

Horwitz was reviewing a book in which an English Marxist, E.P.Thompson, adopted a surprising position—certainly surprising toHorwitz—akin to (2a):

[T]here is a difference between arbitrary power and the rule of law.We ought to expose the shams and inequities which may be con-cealed beneath this law. But the rule of law itself, the imposing ofeffective inhibitions upon power and the defence of the citizen frompower’s all-intrusive claims, seems to me to be an unqualifiedhuman good.24

Thompson’s understanding of the rule of law may be a little moreexpansive than Hart’s and Fuller’s understanding of legality;

23 Morton J. Horwitz, The Rule of Law: An Unqualified Human Good?, 86 YALE L.J.561, 566 (1977) (reviewing DOUGLAS HAY ET AL., ALBION’S FATAL TREE: CRIME AND

SOCIETY IN EIGHTEENTH-CENTURY ENGLAND (1975), and E.P. THOMPSON, WHIGS AND

HUNTERS: THE ORIGIN OF THE BLACK ACT (1975)).24 THOMPSON, supra note 23, at 266.



Thompson seems to adopt a substantive view of the rule of law thatactually identifies it directly with the desirable restraint of power.25

Substantive conceptions of the rule of law are very controversial.26

Fortunately, the idea of a substantive conception of the rule of law isnot at issue in this Article. As we shall see in Part II, it is commonground between our two protagonists—Hart and Fuller—that legalityis best understood in the first instance in formal/structural or proce-dural terms. It is the further moral significance that those formal/structural or procedural principles may have that is at issue in ques-tion (2); to give an affirmative answer to (2) does not mean that onehas adopted a substantive version of the rule of law.

IIHART’S USE OF THE TERM “PRINCIPLES OF LEGALITY”

What did Hart understand the “principles of legality” to mean?Hart’s most extensive discussion of them may be found in a little-known essay—Problems of Philosophy of Law—that he wrote in 1967for Paul Edwards’s Encyclopedia of Philosophy.27

The Encyclopedia essay is divided into three parts: The first partdeals with “Problems of Definition and Analysis,”28 the second with“Problems of Legal Reasoning,”29 and the third with “Problems of theCriticism of Law.”30 After discussing substantive criteria of evalua-tion of legal systems, Hart goes on to say the following:

Laws, however impeccable their content, may be of little service tohuman beings and may cause both injustice and misery unless theygenerally conform to certain requirements which may be broadlytermed procedural . . . . These procedural requirements relate tosuch matters as the generality of rules of law, the clarity with whichthey are phrased, the publicity given to them, the time of theirenactment, and the manner in which they are judicially applied toparticular cases. The requirements that the law, except in special

25 For example, Thompson cites torture and the use of troops against crowds as exam-ples of violations of the rule of law. Id. at 265.

26 See, e.g., B.Z. Tamanaha, The Tension Between Legal Instrumentalism and the Ruleof Law, 33 SYRACUSE J. INT’L L. & COM. 131, 141 (2005) (characterizing Enlightenment ashaving “undermined the very idea that law had or could have any kind of immanentsubstantive content or integrity”); Jeremy Waldron, Legislation and the Rule of Law, 1LEGISPRUDENCE 91, 91 (2007) (arguing that substantive conceptions of rule of law are“needlessly cynical” about capacity of formal/procedural rule of law criteria to constrainpower).

27 H.L.A. Hart, Problems of Philosophy of Law, in 5 THE ENCYCLOPEDIA OF PHILOS-

OPHY 264 (Paul Edwards ed., 1967).28 Id. at 264–68.29 Id. at 268–72.30 Id. at 272–75.



circumstances, should be general (should refer to classes of persons,things, and circumstances, not to individuals or to particularactions); should be free from contradictions, ambiguities, and obscu-rities; should be publicly promulgated and easily accessible; andshould not be retrospective in operation are usually referred to asthe principles of legality. The principles which require courts, inapplying general rules to particular cases, to be without personalinterest in the outcome or other bias and to hear arguments on mat-ters of law and proofs of matters of fact from both sides of a disputeare often referred to as rules of natural justice. These two sets ofprinciples together define the concept of the rule of law . . . .31

The taxonomy may be a little confusing, as there is no universalstandard for parsing these concepts. I would call the principles thatdeal with generality, clarity, prospectivity, and the like formal princi-ples rather than procedural principles; the truly procedural principlesare those to which Hart refers as “principles of natural justice”32 (or,as American lawyers redundantly say, “procedural due process”). Butthe general picture is pretty clear. There are principles about the formthat legal norms should take (formal principles), and there are princi-ples about the broad character of the procedures that should be usedin their application (procedural principles). Together, those principlesof legality and due process add up to what is sometimes called the ruleof law.33

The first set of principles—the formal ones, which Hart called“principles of legality”34—are roughly what Lon Fuller referred to asthe “inner morality of law.”35 Hart’s comments in the Encyclopediaessay—made without any explicit reference to Fuller—are about asclose as Hart ever came to acknowledging the importance of Fuller’scontribution.

Is it of any interest that Hart seems reluctant actually to use (asopposed to mentioning) the term “principles of legality”? Mostly, heattributes its use to others. In the Encyclopedia essay, he speaks of

31 Id. at 273–74.32 For a discussion of the idea of natural justice in this procedural sense, see generally

GEOFFREY A. FLICK, NATURAL JUSTICE: PRINCIPLES AND PRACTICAL APPLICATION (2ded. 1984), and PAUL JACKSON, NATURAL JUSTICE (2d ed. 1979).

33 For important discussions of the rule of law, see FINNIS, supra note 13, at 270–76,JOHN RAWLS, A THEORY OF JUSTICE 235–43 (1971), and RAZ, supra note 14.

34 Hart, supra note 27, at 274.35 FULLER, supra note 17, at 42. Incidentally, I find it odd that Fuller said so little

about the procedural side of the rule of law in chapter two of The Morality of Law, espe-cially in view of his own very intense and focused interest elsewhere in the forms and limitsof adjudication. Id. at 33–94. In fact, Fuller also made the mistake of calling the formalprinciples procedural, as though everything that is not substantive is procedural. For anexample of Fuller’s interest in genuinely procedural issues, see Lon L. Fuller, The Formsand Limits of Adjudication, 92 HARV. L. REV. 353 (1978).



principles that are “usually referred to as the principles of legality,”36

while in The Concept of Law, Hart speaks—all too briefly—of “therequirements of justice which lawyers term principles of legality.”37 Itis not clear why he does this. It is as though he could not bring himselfto use the term in his own voice, for fear that “legality” would connotea more intimate connection between these principles and the veryidea of law than he was comfortable with. In his 1965 review of TheMorality of Law, Hart says that although “principles of legality” isFuller’s term, he (Hart) certainly prefers it to the phrase “the innermorality of law” (for reasons we shall explore in a moment).38

Hart was less coy about the term in his book Law, Liberty andMorality.39 There, he considered the 1961 English decision in Shaw v.Director of Public Prosecutions,40 in which the House of Lordsrevived the old common law offense of conspiracy to corrupt publicmorals. Hart said this about the ruling:

[The House of Lords] seemed willing to pay a high price in terms ofthe sacrifice of other values for the establishment . . . of the Courtsas custos morum. The particular value which they sacrificed is theprinciple of legality which requires criminal offences to be as pre-cisely defined as possible, so that it can be known with reasonablecertainty beforehand what acts are criminal and what are not.41

This is the single exception to Hart’s habit of not using the phrase“principles of legality” himself when he could avoid it. I would notattach any significance to these purely terminological issues, were itnot for Hart’s equivocal response to the substance of the concerns thathe said others had assembled under the auspices of this term.

36 Hart, supra note 27, at 274 (emphasis added).37 HART, supra note 12, at 207 (emphasis added). For a similar locution for “natural

justice”—“the procedural standards known to English and American lawyers as ‘NaturalJustice’”—see id. at 206.

38 H.L.A. Hart, Book Review, 78 HARV. L. REV. 1281, 1284 (1965).39 H.L.A. HART, LAW, LIBERTY, AND MORALITY (1963). This book, Hart’s contribu-

tion to the debate with Lord Patrick Devlin about the use of law to enforce conventionalmorality, was a response to Lord Devlin’s 1958 British Academy Maccabaean Lecture,Morals and the Criminal Law, later reprinted in PATRICK DEVLIN, THE ENFORCEMENT OF

MORALS 1 (1965).40 [1962] A.C. 220 (H.L.) (appeal taken from Eng.) (U.K.). This was the “Ladies’

Directory” case, in which the Law Lords dredged up a nonstatutory common law crime ofconspiracy to corrupt public morals and used it to convict a man who had published adirectory advertising the services of prostitutes. See id. at 266–68.

41 HART, supra note 39, at 12.



IIIWHAT DOES HART SAY ABOUT LEGALITY?

What, apart from taxonomy and terminology, did Hart actuallysay about the principles variously referred to as principles of legalityor the rule of law? “Rather little” is the answer, and though much ofwhat he did say is suggestive, very little of it is consistently presentedor rigorous by the standards of other aspects of Hart’s legal philos-ophy. In this Part, I will review the record of Hart’s comments invarious writings on this matter. In Part IV, I will substantiate my com-plaint about its inconsistency.

A. Hart’s 1958 Holmes Lecture

In his Holmes Lecture, Hart first mentioned the principles oflegality in an effort to defend Jeremy Bentham against the charge ofindifference to the evaluation of law:

One by one in Bentham’s works you can identify the elements ofthe Rechtstaat and all the principles for the defense of which theterminology of natural law has in our day been revived. Here areliberty of speech, and of press, the right of association, the need thatlaws should be published and made widely known before they areenforced, the need to control administrative agencies, the insistencethat there should be no criminal liability without fault, and theimportance of the principle of legality, nulla poena sine lege.42

The most striking thing about this paragraph is the way that it runstogether, in a rather casual way, a variety of ideals. Some of them arepolitical ideals, such as free speech and freedom of assembly, sepa-rable and distinct on almost any account from the idea of law; some,like the requirement that laws be published and the principle nullapoena sine lege (no punishment without law), are arguably (thoughnot indisputably) connected to the concept of law; and some, like con-trol of administrative agencies and the principle of no liability withoutfault, fall in between.

Hart links all of these and uses them, along with some observa-tions about Bentham’s opposition to slavery, as evidence for the gen-eral proposition that Bentham was not a “dry analyst[ ] fiddling withverbal distinctions while cities burned, but . . . the vanguard of amovement which laboured with passionate intensity and much successto bring about a better society and better laws.”43 The implication ofrunning all these together is that the principles of legality—as much asthe principle of free speech or opposition to slavery—are simply

42 Hart, supra note 1, at 595–96.43 Id. at 596.



moral criteria for better laws (i.e., what the law ought to be) and thatthey have little or nothing to do with the concept of law itself. Theprinciples of legality are all associated without distinction in this pas-sage with “principles for the defense of which the terminology of nat-ural law has in our day been revived.”44 And the assortment ofprinciples that Hart assembled under this heading reinforces Fuller’sobservation that when law is distinguished from morality in positivistjurisprudence, “the word ‘morality’ stands indiscriminately for almostevery conceivable standard by which human conduct may be judgedthat is not itself law.”45

Another point in the 1958 lecture at which Hart touched on theseissues was his discussion of Nazi rule in Germany. Hart alluded to theprinciples of legality in a suggestion he made concerning the GrudgeInformer Case, in which the question arose of punishing a woman in1949 for having denounced her husband to the authorities in 1944under oppressive Nazi statutes.46 Hart took the position that, insteadof declaring the Nazi statutes to have been nullities, it would havebeen better to enact a statute after 1945 and apply that statute retro-actively in order to punish the woman’s vindictive action:

Odious as retrospective criminal legislation and punishment may be,to have pursued it openly in this case would at least have had themerits of candour. It would have made plain that in punishing thewoman a choice had to be made between two evils, that of leavingher unpunished and that of sacrificing a very precious principle ofmorality endorsed by most legal systems.47

Here, the principle of prospectivity is treated as just another moralprinciple, albeit “a very precious principle of morality.” It is“endorsed by most legal systems” but is not spoken of as being tied inany special way to the concept of law.

Principles of legality are sometimes thought to include therequirement that laws be general rather than in personam or ad hoc.Hart also addressed the significance of this requirement in his 1958lecture:

If we attach to a legal system the minimum meaning that it mustconsist of general rules—general both in the sense that they refer tocourses of action, not single actions, and to multiplicities of men,not single individuals—this meaning connotes the principle of

44 Id. at 595.45 Fuller, supra note 2, at 635.46 Fuller gives a good account of this case, id. at 652–57, setting out the statutes, id. at

653, 654. Hart’s discussion of the case is in Positivism and the Separation of Law andMorals. See Hart, supra note 1, at 618–21; see also David Dyzenhaus, The GrudgeInformer Revisited, 83 N.Y.U. L. REV. 1000 (2008).

47 Hart, supra note 1, at 619.



treating like cases alike, though the criteria of when cases are alikewill be, so far, only the general elements specified in the rules. It is,however, true that one essential element of the concept of justice isthe principle of treating like cases alike. This is justice in the admin-istration of the law, not justice of the law. So there is, in the verynotion of law consisting of general rules, something which preventsus from treating it as if morally it is utterly neutral, without anynecessary contact with moral principles.48

In this passage, generality (as a principle of legality) is associatedexplicitly with the concept law or at any rate with the concept of alegal system. It is part, Hart says, of the “minimum meaning” of thelatter, and it does have some limited moral significance. This state-ment is about as close as Hart ever came to considering, together andconsistently, the two questions that I said earlier Fuller had put on theagenda of jurisprudence.

It is interesting, finally, that in the same passage Hart also associ-ated generality with his discussion of what he referred to as “naturaljustice” (or what we might call procedural due process) as a moral orquasi-moral principle related to the concept of a legal system:

Natural procedural justice consists . . . of those principles of objec-tivity and impartiality in the administration of the law which imple-ment just this aspect of law [treating like cases alike] and which aredesigned to ensure that rules are applied only to what are genuinelycases of the rule or at least to minimize the risks of inequalities inthis sense.49

This may underestimate somewhat the tasks of procedural due pro-cess, which go beyond ensuring the consistency that genuine gener-ality requires. Still, it is, for Hart, an important early concession.

B. Hart’s Book: The Concept of Law

In many respects, Hart’s 1958 lecture adumbrated more extensivearguments about generality and its connection to due process asaddressed in his magisterial 1961 work on jurisprudence, The Conceptof Law.50 Chapter VIII of The Concept of Law is devoted to a discus-sion of various aspects and meanings of morality (which Hart’s posi-tivism strives to distinguish from law);51 the section most relevant hereis devoted specifically to justice. Hart argues that justice—as a dis-

48 Id. at 623–24.49 Id. at 624.50 HART, supra note 12.51 In The Concept of Law, Hart provides the following definition of legal positivism:

“Here we shall take Legal Positivism to mean the simple contention that it is in no sense anecessary truth that laws reproduce or satisfy certain demands of morality . . . .” Id. at185–86.



tinctive segment of morality—often involves ideas of equality andtreating similar cases in a similar way:

[I]t might be said that to apply a law justly to different cases issimply to take seriously the assertion that what is to be applied indifferent cases is the same general rule, without prejudice, interest,or caprice. This close connection between justice in the administra-tion of the law and the very notion of a rule has tempted somefamous thinkers to identify justice with conformity to law. Yetplainly this is an error . . . for such an account of justice leaves unex-plained the fact that criticism in the name of justice is not confinedto the administration of the law in particular cases, but the lawsthemselves are often criticized as just or unjust.52

Hart goes on to argue, fairly convincingly, that this more critical use ofjustice cannot be accounted for in terms of any background idea thatjustice (in a more limited sense) and generality, as a legalistic ideal,might share.53

Legality is sometimes associated with the twin principles that lawsshould be clear and that it should be clear which norms have the statusof law and which norms do not. Hart discusses both of these ideas inThe Concept of Law and generally takes the position that, althoughthey are important principles, they are not the be-all and end-all oflegal morality. So far as primary rules (i.e., rules governing basic con-duct) are concerned, he maintains that we have to balance the needfor certainty—“the need for certain rules which can, over great areasof conduct, safely be applied by private individuals to themselveswithout fresh official guidance”—against the need to leave certainissues open so that they can be settled as they arise in concrete cases.54

The reference to safety in regard to the principle of certainty is keyhere: Many defenders of the rule of law have emphasized exactly thissort of safety.55

Hart discusses the issue of clarity with regard to the recognitionof law—clarity as to which norms are law and which are not—in the“Postscript” to The Concept of Law, added by the editors of thatbook’s second edition in 1994.56 Hart had originally argued that the

52 Id. at 161. Note that “some famous thinkers” is not an allusion to Lon Fuller butrather to some obscure views of Thomas Hobbes and John Austin. Id. at 299–300 n.161.See also id. at 206 (briefly discussing generality as “the germ at least of justice”).

53 Id. at 161–67.54 Id. at 128–30 (“[W]e should not cherish, even as an ideal, the conception of a rule so

detailed that the question whether it applied or not to a particular case was always settledin advance.”).

55 For a recent example, consider the use of the image of law as a safe causeway inJeffrey Kahn, The Search for the Rule of Law in Russia, 37 GEO. J. INT’L L. 353, 354 (2006).

56 HART, supra note 12, at viii–x, 238–76.



rule of recognition served the need of every system of social controlfor “certainty” in people’s understanding of which rules are going tobe coercively enforced by society with the centralized social and phys-ical pressure that it coordinates.57 But, in the “Postscript,” Hartargued once again that the need for certainty is not an absolute. Heassociated this playing-down of certainty with his willingness, at theend of his life, to have his work characterized as a form of “softpositivism”:

It is of course true that an important function of the rule of recogni-tion is to promote the certainty with which the law may be ascer-tained. . . . But the exclusion of all uncertainty at whatever costs inother values is not a goal which I have ever envisaged for the rule ofrecognition. . . . Only if the certainty-providing function of the ruleof recognition is treated as paramount and overriding could theform of soft positivism that includes among the criteria of law con-formity with moral principles or values which may be controversialbe regarded as inconsistent.58

Apart from this argument about generality and uncertainty,Hart’s most important discussion of legality in The Concept of Law isalso his briefest. In the course of a discussion of the various ways inwhich law might be related to morality, Hart invites us to consider

what is in fact involved in any method of social control . . . whichconsists primarily of general standards of conduct communicated toclasses of persons, who are then expected to understand and con-form to the rules without further official direction. If social controlof this sort is to function, the rules must satisfy certain conditions:they must be intelligible and within the capacity of most to obey,

57 See id. at 94 (“The simplest form of remedy for the uncertainty of the regime ofprimary rules is the introduction of what we shall call a ‘rule of recognition.’”).

58 Id. at 251–52. I regard Hart’s willingness to entertain soft positivism as an issuemostly distinct from the implications of his equivocal views about legality. I will develop arelated analogy at the end of this Article. See infra note 116 and accompanying text. But itis interesting that he sees a possible negative connection: The more one emphasizes princi-ples of legality requiring certainty in the law, the more one might be driven toward aharder form of positivism. I have sometimes wondered whether hard positivism might notbe associated naturally with normative positivism—that is, with views like ThomasHobbes’s that give substantive moral reasons for requiring a clear distinction between lawand morality. See Jeremy Waldron, Normative (or Ethical) Positivism, in HART’S POST-

SCRIPT: ESSAYS ON THE POSTSCRIPT TO THE CONCEPT OF LAW 411, 412–13 (Jules L.Coleman ed., 2001); THOMAS HOBBES, LEVIATHAN 183–200 (Richard Tuck ed., CambridgeUniv. Press 2000) (1651) (asserting that law’s function—to promote peace by settling orpreempting moral disputes—would be undermined by need to engage in tendentious mor-alizing in order to identify law). Hart raises the alternative possibility that, leaving asidesubstantive values like peace, hard positivism is more responsive to formal and proceduralprinciples of legality such as the principle of certainty.



and in general they must not be retrospective, though exceptionallythey may be.59

Hart said that “[p]lainly these features of control by rule are closelyrelated to the requirements of justice which lawyers term principles oflegality.”60 (Again, note the oratio obliqua use of that phrase.) But,in The Concept of Law, Hart had no further interest in explicatingwhat “principles of legality” might mean. His main interest seemed tobe squelching any inference that might lead from an acknowledgmentof the value of the principles of legality to a position like Fuller’sabout law’s overall moral potential. Hart alluded briefly to Fuller’sown view of these principles—“one critic of positivism has seen inthese aspects of control by rules, something amounting to a necessaryconnection between law and morality”61—but made no comment ofhis own except to say, acidly, that these formal requirements are“unfortunately compatible with very great iniquity.”62 That, heseemed to indicate, was more or less all that needed to be said. In hismind, it seemed that the crucial thing was to protect the separabilitythesis.63 According to Hart, Fuller was wrong in thinking that theimposition of the requirements of legality guaranteed that the lawswould not be evil. Once that was established, the principles of legalitywere of little interest for jurisprudence.

C. Hart’s Encyclopedia Essay: Problems of Philosophy of Law

In his essay for Paul Edwards’s Encyclopedia of Philosophy, Hartagain raised the point that the principles of legality were “compatiblewith very great iniquity.” Even if principles of legality are very impor-tant, he said, we must not infer that “it will always be reasonable ormorally obligatory for a man to obey the law when the legal systemprovides him with [the] benefits [of principles of legality and naturaljustice], for in other ways the system may be iniquitous.”64 This issomewhat less dismissive than his comments in The Concept of Law.For one thing, it acknowledges that there are important moral valuesunderpinning the principles of legality (even though they may be out-weighed by this ubiquitous “iniquity” that he harps so heavily upon).Later in this Article, we will have to address the difference betweensaying (i) that nothing of moral significance follows from the fact thata system conforms (or fails to conform) to principles of legality, and

59 HART, supra note 12, at 206–07.60 Id. at 207.61 Id.62 Id.63 See supra note 51 for Hart’s clearest summary of the separability thesis.64 Hart, supra note 27, at 274.



(ii) that nothing of conclusive moral significance follows from the factthat a system conforms (or fails to conform) to principles of legality.65

In the Encyclopedia essay, Hart clearly seems to be saying (ii),whereas it is not clear which of these positions he was asserting in TheConcept of Law.

Even if Hart acknowledges that the principles of legality haveaffirmative moral significance, such significance still may have manydimensions. We still must ask how much conformity to the principlesof legality affects the substantive value or justice of the rules, and howmuch the absence of such conformity contributes to substantive injus-tice and iniquity in the law.

Another dimension, mentioned by Hart in the passage justquoted, is our obligation to obey the laws. Hart says that we must notinfer anything about political obligation from the fact that laws con-form to legality.66 Again, this may be for two reasons, partly analo-gous to (i) and (ii) in the previous paragraph. It may be wrong toinfer obligation from legality, (i*), because considerations of legalitymake no contribution at all to the question of our moral obligation toobey the laws (whether or not they make any contribution to thequestion of the substantive justice of the norms) or (ii*) because,although considerations of legality make some contribution to thequestion of our moral obligation to obey the laws, they do not con-clude or settle that question. Hart’s account prompts us to make thesedistinctions, though Hart himself shows little interest in exploringthem.

In the Encyclopedia essay, Hart also talks about the efficiencyimplications of the principles of legality. He says that “general rulesclearly framed and publicly promulgated are the most efficient formof social control.”67 He is describing efficiency from the point of viewof the ruler. But then he immediately goes on to say that “from thepoint of view of the individual citizen, they are more than that.”68

Conformity with these principles is “required if [the individual citizen]is to have the advantage of knowing in advance the ways in which hisliberty will be restricted in the various situations in which he may findhimself, and he needs this knowledge if he is to plan his life.”69

Hart also says that generality helps the individual citizen becauseit gives him information about what others will be held to, which“increases the confidence with which he can predict and plan his

65 See infra Part V.66 Hart, supra note 27, at 275.67 Id. at 274.68 Id.69 Id.



future.”70 Efficiency may still be involved here, but it is no longerefficiency from the point of view of the ruler. It is respect by the rulerfor the conditions of efficiency considered from the point of view ofthe citizen. Insistence upon such respect is almost certainly going tobe a moral requirement, for it asks the ruler to facilitate the pursuit ofinterests other than his own even when doing so conflicts with his owninterests.

D. Hart’s Review of The Morality of Law

This distinction between the different ways in which efficiencymight feature in lawmaking may also help in our assessment of a com-ment that Hart made in his review of Fuller’s book: Hart said thatFuller’s “insistence on classifying these principles of legality as a‘morality’ is a source of confusion both for him and his readers.”71

They may, said Hart, be “principles of good legal craftsmanship,”72

but that does not make them into a morality. In a vivid analogy, Hartasked whether a set of craft principles for poisoners would qualify as a“morality” of poisoning:

Poisoning is no doubt a purposive activity, and reflections on itspurpose may show that it has its internal principles. (“Avoid poisonshowever lethal if they cause the victim to vomit,” or “Avoid poisonshowever lethal if their shape, color, or size is likely to attractnotice.”) But to call these principles of the poisoner’s art “themorality of poisoning” would simply blur the distinction betweenthe notion of efficiency for a purpose and those final judgmentsabout activities and purposes with which morality in its variousforms is concerned.73

Now, this analogy depends on seeing the principles of legalitysimply as principles for the efficient pursuit of the lawmaker’s pur-pose, and some of what Fuller says in his analogy about craftsmanshipand carpentry does encourage that reading.74 No doubt, if the princi-ples of legality were just principles of efficiency from the lawmaker’spoint of view—i.e., “How To Do Things with Rules”—then it would

70 Id.71 Hart, supra note 38, at 1285.72 Id. at 1286.73 Id.74 See FULLER, supra note 17, at 96, 155–56. Fuller also presents these ideas in Posi-

tivism and Fidelity to Law, supra note 2, at 644–45, and encourages this instrumentalistview. But in the Reply to Critics added to the revised edition of The Morality of Law,Fuller adopted a much more nuanced position in which he indicated that the most impor-tant requirements of legality should not be viewed in this way and attacked his critics for“maintaining the view that the principles of legality represented nothing more than maximsof efficiency for the attainment of governmental aims.” FULLER, supra note 17, at 214. Seegenerally id. at 200–24.



be inappropriate to call them a morality, at least (as Hart said)without showing that “the purpose of subjecting human conduct to thegovernance of rules, no matter what their content” was “of ultimatevalue in the conduct of life.”75

But later in the review, Hart seems to equivocate in an attempt torespond more defensively to Fuller’s characterizations of legal posi-tivism. Fuller suggested that legal positivism was insufficiently con-cerned with legality—that positivists (in Hart’s words) “‘cannot evenexplain what would be wrong with a system of laws that were whollyretroactive,’ and that we cannot give any adequate explanation of whynormally legal rules are general.”76 In response, Hart contends that itis perfectly possible to develop an account of what would be wrongwith such a system. And, though he does not say so explicitly, he inti-mates that such an account would be firmly located within the realmof the moral. His response takes us back to one of the earliestpassages we considered in the Holmes Lecture:77

Why, to take the simplest instances, could not writers like Benthamand Austin, who defined law as commands, have objected to asystem of laws that were wholly retroactive on the ground that itcould make no contribution to human happiness and so far as itresulted in punishments would inflict useless misery? Why shouldnot Kelsen or I, myself, who think law may be profitably viewed as asystem of rules, not also explain that the normal generality of law isdesirable not only for reasons of economy but because it will enableindividuals to predict the future and that this is a powerful contribu-tion to human liberty and happiness?78

Hart goes on to imagine that Fuller might complain that Hart’sown use or Bentham’s use of these principles is not really moral,because it is just oriented to human happiness; Hart wrote that theprinciples of legality “are valued so far only as they contribute tohuman happiness or other substantive moral aims of the law,”79 andFuller might characterize this position as implying that those princi-ples are not really moral in themselves. Hart regards that quite rightlyas a distinction without a difference.80 But given Hart’s insistencehere on the moral character of his own use of principles of prospec-tivity and generality in the evaluation of law (or Bentham’s, orKelsen’s), it is now difficult to see why Hart thought himself entitled

75 Hart, supra note 38, at 1287.76 Id. at 1290 (quoting FULLER, supra note 17, at 147).77 See supra note 42 and accompanying text.78 Hart, supra note 38, at 1290–91.79 Id. at 1291.80 Id.



to say that Fuller’s insistence on the moral character of these princi-ples is like talking about the morality of poisoning.

E. Hart’s Book: Punishment and Responsibility

Apart from the passage on Shaw v. Director of Public Prosecu-tions that I cited earlier,81 the only other significant mention of (orallusion to) the principles of legality in Hart’s work is in his book onthe jurisprudence and policy of the criminal law, Punishment andResponsibility.82 In an important passage in that book, Hart consid-ered what would happen if Barbara Wootton’s proposal—that strictliability should comprehensively replace ordinary criminal liability—were to prevail:83

Among other things, we should lose the ability which the presentsystem in some degree guarantees to us, to predict and plan thefuture course of our lives within the coercive framework of the law.For the system which makes liability to the law’s sanctions depen-dent upon a voluntary act not only maximizes the power of the indi-vidual to determine by his choice his future fate; it also maximizeshis power to identify in advance the space which will be left open tohim free from the law’s interference.84

Hart presented the same point in a more sustained way in a 1965review of Wootton’s book:

In a system in which proof of mens rea was no longer a necessarycondition for conviction the occasions for official interferences inour lives would be vastly increased. . . . [E]very blow, even if it wasapparent that it was accidental or merely careless . . . would in prin-ciple be a matter for investigation under the new scheme. Thisexpansion of police powers would bring with it great uncertainty forthe individual citizen and, though official interference with his lifewould be more frequent, he will be less able to predict their inci-dence if any accidental breach of the criminal law may be an occa-sion for them.85

Hart here makes no reference, explicit or implicit, to the phrase “prin-ciples of legality.” But it is clear that he is invoking such principlesand indeed values and concerns traditionally associated with the rule

81 See supra notes 40–41 and accompanying text.82 H.L.A. HART, PUNISHMENT AND RESPONSIBILITY: ESSAYS IN THE PHILOSOPHY OF

LAW (1968).83 For Baroness Wootton’s views, see generally BARBARA WOOTTON, CRIME AND THE

CRIMINAL LAW (2d ed. 1981).84 HART, supra note 82, at 181–82. For bringing this passage to my attention, I am

grateful to the excellent discussion in Hamish Stewart, Legality and Morality in H.L.A.Hart’s Theory of Criminal Law, 52 SMU L. REV. 201, 204 (1999).

85 H.L.A. Hart, Book Review, 74 YALE L.J. 1325, 1330 (1965) (reviewing BARBARA

WOOTTON, CRIME AND THE CRIMINAL LAW (1963)).



of law. And he shows no reluctance to say that these can be used asgenuine policy-based or moral critiques of legislative proposals; thatis, they are not just instrumentalist critiques to the effect that alawmaker who set up a Wootton-type scheme would be frustrating hisown purposes.

* * * *

To my knowledge, what I have set out in this Part is the sum totalof H.L.A. Hart’s observations on legality, natural justice (proceduraldue process), and the rule of law—and their relation to morality, onthe one hand, and the concepts law and legal system, on the other. Inthe next Part, I will analyze the consistency of all of this.

IVHART’S INCONSISTENCY

I implied at the outset that Hart’s discussion of legality is equiv-ocal.86 It often seems to be motivated by a desire to say nothing morethan is necessary to dispatch Lon Fuller’s critique—and, if what is nec-essary to refute Fuller in one discussion is inconsistent with what isnecessary to refute Fuller in another context, Hart seems to rest hishopes of prevailing on the fact that many of his readers will be moreinterested in Fuller’s discomfiture than in the (in)consistency of Hart’srefutation. It is, I think, a shabby episode in the history of modernpositivist legal philosophy—the more so since Hart, even according toFuller, is supposed to hold the high ground as far as standards of ana-lytic clarity are concerned.87 Hart’s treatment of Fuller gives stan-dards of analytic clarity in legal philosophy a bad name.

The basic contradiction in Hart’s account of the principles oflegality lies in the answers he gives to the two questions that I said atthe outset were suggested by Fuller’s reflections on the case of NaziGermany:

(1) Is observance of the principles of legality among the necessarycriteria for the application of the concepts law and legal system to asystem of rule?; and(2) Does observance of the principles of legality make an affirma-tive difference to the moral quality of a system of rule?

To question (1) Hart says two things, corresponding roughly tothe possible answers (1a and 1b) we imagined in Part I. Following the

86 See supra text accompanying note 9.87 See Fuller, supra note 2, at 630 (“Professor Hart has made an enduring contribution

to the literature of legal philosophy. I doubt if the issues he discusses will ever againassume quite the form they had before being touched by his analytical powers.”).



reasoning of (1b), he suggests that principles of legality are to begrouped among those principles and values clearly distinguished fromthe concept of law by the positivist’s separability thesis. They are notcriteria for calling something law or a legal system. Instead, they areto be grouped among “the principles for the defense of which the ter-minology of natural law has in our day been revived . . . . [like] libertyof speech, and of press” and also like the principle forbiddingslavery.88

But Hart’s position also sounds like (1a) when he contends thatthe principles of legality are in fact very closely related to the ideas oflaw and a legal system. He says of the principle of generality, forexample, that “we attach to a legal system the minimum meaning thatit must consist of general rules.”89 Moreover, toward the end of TheConcept of Law, he associates Fullerian principles with the viability ofany efficacious method of social control.90 That last point may notexactly involve a logical connection to the morality of the method ofsocial control,91 but it does seem to resemble a fundamental criterialconnection of the kind that makes up the tissue of other aspects ofHart’s jurisprudence (such as the connection between law and theidea of secondary rules or the specification of the two existence condi-tions for a legal system).92

The apparent inconsistency of Hart’s responses to the first ques-tion, resembling both (1a) and (1b), might be resolved by reference tothe carelessness of Hart’s formulations, were it not for the fact that itmatches an inconsistency in the way he answers the second question.For, in response to question (2)—which asks about the relationbetween the principles of legality and moral values like justice, rights,and the common good—Hart also returns two quite different answers.Following (2a), on the one hand, he groups the principles of legalityamong the criteria of substantive justice. We saw this in his 1958 dis-cussion of Bentham,93 in his treatment of the Grudge Informer Case,94

and in his discussion of Shaw v. Director of Public Prosecutions.95

88 See Hart, supra note 1, at 595–96.89 Id. at 623.90 See supra notes 59–60 and accompanying text.91 But as the controversy about Dworkin’s “semantic sting” has shown, logical connec-

tions are not really the issue. See RONALD DWORKIN, LAW’S EMPIRE 45–46 (1986) (char-acterizing positivist theories as “semantic” theories that cannot accommodate reality oftheoretical disagreement about law); HART, supra note 12, at 244–48 (refuting Dworkin’scharacterization and asserting openness of positivism to such disagreement).

92 See HART, supra note 12, at 91–99 (discussing secondary rules); id. at 110–17 (dis-cussing existence conditions).

93 Hart, supra note 1, at 594–600; see also supra note 42 and accompanying text.94 Hart, supra note 1, at 618–20; see also supra notes 46–47 and accompanying text.95 HART, supra note 39, at 12; see also supra notes 39–41 and accompanying text.



Recall that in the Grudge Informer Case, he referred to the principleprohibiting retroactive laws as “a very precious principle of moralityendorsed by most legal systems.”96 This is not just idle talk. Hartgives a brief but substantive account of why the principles of legalityhave moral value. He says, in his review of Fuller’s book, that laws“that were wholly retroactive . . . could make no contribution tohuman happiness and so far as [they] resulted in punishments wouldinflict useless misery . . . .”97 And in his review of Baroness Wootton’sproposal, he suggests that the element of fair warning and predict-ability that clear, published, and prospective laws require is beneficialto citizens’ ability to exercise freedom and plan for the future.98

On the other hand, when he is confronting Fuller’s claim that weshould pay particular attention to the moral significance of legality,Hart beats a hasty retreat from such characterizations. In those con-texts, his position follows (2b): He denies that that the principles oflegality have any particular moral significance. He says not only thattheir observance is “compatible with very great iniquity”99 but alsothat he is utterly puzzled about why Fuller would refer to them usingthe term “morality.”100

It is a distressing picture. Hart’s honest inclination seems to be toanswer “yes” to both of our questions, at least when they are posedseparately. He really does seem to acknowledge a criterial connectionbetween the idea of a legal system and at least some of the principlesof legality, most evidently in his remarks on generality. Additionally,he seems to want to insist that when Bentham and others applied prin-ciples of legality to the evaluation of law, they were applying criteriathat had genuine moral significance.

However, the combination of these two positions—(1a) and(2a)—looks likely to cause problems for the distinctive positivistthesis of the separability of law and morality, because it implies thatone of the criteria for calling something a legal system has genuinemoral significance. So Hart takes care to separate his assent to ques-tion (1) from his assent to question (2). The assents tend to occur indifferent writings; when they occur in the same article (as in thereview of Fuller’s book), he makes sure that there are a few pagesseparating them.101 That way, he can give the impression that when

96 Hart, supra note 1, at 619; see also supra note 47 and accompanying text.97 Hart, supra note 38, at 1291; see also supra note 78 and accompanying text.98 Hart, supra note 85, at 1330.99 HART, supra note 12, at 207.

100 Hart, supra note 38, at 1286–87.101 Id. at 1286–87, 1290–91.



he assents to question (2), he is conjoining this with a negative answerto question (1), and vice versa.

But for anyone who is willing and able to sustain their attentionfrom page to page or from article to article, Hart’s position is all overthe place. In any given instance, he seems to be combining his answerto one of our questions with whatever answer to the other question isnecessary to make Fuller’s position look untenable. Because Fuller’sposition has two parts, Hart achieves this by variously maintaining allfour of the possible answers that we have outlined, hoping that we donot notice that two of them, when taken in combination, are incom-patible with positivism as he seems to understand it.

FIGURE 1(1) Is observance of the principles of legalityamong the necessary criteria for the application ofthe concepts law and legal system to a system ofrule?

(1a) YES (1b) NO(2) Does observance of theprinciples of legality make an (2a) YES (NOT OK w/ positivism) (OK w/ positivism)affirmative difference to themoral quality of a system of (2b) NO (OK w/ positivism) (OK w/ positivism)rule?

The position looks like Figure 1 above. Hart answers “yes” and“no” to both of the questions but wants us, at any given time, to focusonly on the combination of “yes” to question (1) and “no” to question(2) or on the combination of “no” to question (1) and “yes” to ques-tion (2). Both combinations are acceptable from a legal positivistpoint of view. Negative answers to both questions would also beacceptable for a positivist, but affirmative answers to both would not.Yet Hart seems to have committed himself, over time, to each of theseoptions (including “yes” to question (1) and “yes” to question (2),which is not a combination that is acceptable to most positivists).

VPLEAS IN MITIGATION

A. Reconciling Opposite Answers

I have said some hard things about Hart’s inconsistency. It istime now for pleas in mitigation. The contradictions in Hart’s positionwould be mitigated somewhat if we could somehow reconcile (1a)with (1b), and/or (2a) with (2b).

On question (1)—asking if there is a criterial connection betweenprinciples of legality and the concepts law and legal system—therecertainly does seem to be some room in between a strongly affirma-



tive and a flatly negative answer. For one thing, the principles oflegality are several in number, and although Hart is willing to say thatlaw might logically connote generality,102 he may not be willing to sayanything like that for all or even any of the other principles of legality.Furthermore, even Fuller acknowledges that the criterial relationbetween law and legal system, on the one hand, and the principles ofthe inner morality of law that he identifies, on the other, is quite loose.For example, although Fuller suggests that general and widespreaduse of retroactive decrees would undermine the claim of a system ofrule to be called rule by law, he acknowledges that “in England andAmerica it would never occur to anyone to say that ‘it is in the natureof law that it cannot be retroactive.’”103 He says something similarabout secrecy: An inadvertent failure to publish some set of regula-tions may not be incompatible with a system’s claims to be a system ofrule by law, but widespread and deliberate use of secret decrees maybe.104

Clearly there is some room for looseness here, and one couldimagine an honorable attempt to rescue Hart’s position from theclutches of inconsistency by arguing that when he said (1a)—thatlegality did have a criterial relation to law—he imagined a very looserelation, and that when he said (1b)—that legality did not have acriterial relation to law—he meant only that a simple failure oflegality did not result in an immediate failure of the application ofother legal predicates. Hart himself, however, did not think these pos-sibilities worth exploring in relation to the various positions he heldconcerning the answer to question (2).

As for the possibility of reconciling Hart’s opposite answers toquestion (2), I have said a number of times that a person who believesthat legality has moral significance need not also believe that legalityconclusively determines the issue of the moral quality of a givenlaw.105 Certainly, he need not believe that legality conclusively deter-mines the political obligation implied by such a law.

Thus, while Hart does seem to believe that conformity to theprinciples of legality is compatible with great iniquity, this claim couldsurvive an affirmative answer to (2)—that legality does have moralsignificance—if, for example, that significance was just one among anumber of factors that might enter into a law’s overall moral value.Indeed, this position would be consistent with quite a strong versionof (2a): We might say that conformity to the principles of legality

102 Hart, supra note 1, at 595–96.103 Fuller, supra note 2, at 650.104 Id. at 651.105 See supra Part I (refining and discussing question (2)).



always makes things better, even though it is not necessarily capableof rescuing a law from the gross iniquity of its content. On thisaccount, we might say, for example, that even if Nazi rule did not sat-isfy the principles of legality, other forms of iniquitous rule might doso—for example, South African apartheid law or antebellumAmerican slave law. In those cases, iniquitous legal systems wouldhave been even worse had the principles of legality not beenobserved. That is one possible way of rescuing Hart from inconsis-tency. It requires paying attention to the plurality of considerationsthat enter into the overall evaluation of any law or legal system.

There exists an even more intriguing possibility. We might saythat, quite apart from other substantive moral factors, legality itselfmight work in two directions. On the one hand, conformity to theprinciples of legality does tend to mitigate certain aspects of injusticethat might otherwise be present, even if it fails to entirely redeem thelaw in question. On the other hand, that very same conformity to theprinciple of legality might also have the potential in some cases toaggravate injustice. The same factor—legality—may work both ways.Hart comes close to saying this in the section of The Concept of Lawin which he deals most explicitly with the separability thesis. He saysthat our reflections on the role of secondary rules in improving law’scertainty and knowability bring us face to face with “a sobering truth”:

[T]he step from the simple form of society, where primary rules ofobligation are the only means of social control, into the legal worldwith its centrally organized legislature, courts, officials, and sanc-tions brings its solid gains at a certain cost. The gains are those ofadaptability to change, certainty, and efficiency, and these areimmense; the cost is the risk that the centrally organized power maywell be used for the oppression of numbers with whose support itcan dispense, in a way that the simpler regime of primary rulescould not.106

The very efficiency and centralization that give people their assurancethat they know where they stand vis-a-vis the law also give the statethe means to oppress and exploit them more effectively than wouldotherwise be the case. Hart goes on to say that it is “[b]ecause thisrisk has materialized and may do so again”107 that we should be verywary of any attempt to show that law as such is necessarily moral.Notice how different this is from any analytic version of the separa-bility thesis. The claim now is that legality contributes to the moralquality of the law in some ways and detracts from the moral quality of

106 HART, supra note 12, at 202.107 Id. I have discussed this also in Jeremy Waldron, All We Like Sheep, 12 CANADIAN

J.L. & JURISPRUDENCE 169, 174–75, 179 (1999).



the law in other (often very similar) ways—and there is no telling howthings will fall out overall. This is a very complex denial (or combina-tion of denials) of the separability thesis, not a version of it.108

There remains one other possibility for reconciling Hart’s incon-sistencies. We noticed earlier that some of Hart’s denials of the moralsignificance of legality are not about the moral quality of the laws butabout the presence or absence of a link between legality and politicalobligation.109 We might easily reconcile a denial that legality gener-ates political obligation with an acknowledgment that legality makesan affirmative difference to the overall moral quality of a law. Forexample, a consent theorist of political obligation would believe that,unless consent is present, the moral quality of a law is irrelevant to thepolitical obligation to obey it.110 It could be the most just law imagi-nable (and legality might have made a considerable contribution to itsjustice), but the consent theorist would still require consent for obliga-tion. So that is another way of reconciling a version of (2a) with aversion of (2b).

As a matter of fact, however, few political theorists take the viewthat obligation is wholly detached from the question of justice in thisway. Some follow John Rawls and say that we have an obligation tosupport just institutions and laws precisely because they are just.111 Inthat case, the contribution of legality to justice (whatever it may be)could not be separated from the obligation aspect in the way that con-sent theory indicates, though it might still be true that legality alone

108 I owe this point to Tom Campbell, Professor and former Dean of Law at theAustralian National University, in discussion from the floor during a legal theory confer-ence at the University of Sydney in September 2002.

109 See Hart, supra note 27, at 274–75 (noting that adherence to principles of legalitymay not compel obligation to obey law where legal system is otherwise iniquitous); see alsosupra notes 64–66 and accompanying text (discussing Hart’s argument that conforming tolegality does not guarantee obligation to obey laws). Consider also the way Hart runsthese issues together in the following passage (addressed to Gustav Radbruch’s repudia-tion of positivism):

[E]verything that [Radbruch] says is really dependent upon an enormous over-valuation of the importance of the bare fact that a rule may be said to be avalid rule of law, as if this, once declared, was conclusive of the final moralquestion: “Ought this rule of law to be obeyed?”

Hart, supra note 1, at 618; see also Hart, supra note 27, at 275 (“It seems clear that themere existence of a legal system, irrespective of the character of its laws, is not sufficient inany intelligible theory of morality to establish that a person ought morally to do what itslaws require him to do.”).

110 For the consent theory of political obligation, see A. JOHN SIMMONS, MORAL PRINCI-

PLES AND POLITICAL OBLIGATIONS 57–74 (1979).111 See RAWLS, supra note 33, at 115 (“[T]he duty of justice. . . . requires us to support

and to comply with just institutions that exist and apply to us.”); Jeremy Waldron, SpecialTies and Natural Duties, 22 PHIL. & PUB. AFF. 3, 3 (1993) (defending and elaboratingtheory “that we have a natural duty to support the laws and institutions of a just state”).



does not determine the issue of justice to which Rawls’s theory ofpolitical obligation is responsive. Others have toyed with principles ofintegrity or fairness as the basis of political obligation—take RonaldDworkin’s theory of integrity, for example, or Hart’s own commit-ment to the principle of fair play.112 These accounts of political obli-gation are not bound to the justice of a law as tightly as Rawls’s theoryis, but neither are they wholly detached from it. Some aspects of jus-tice are certainly relevant to integrity and to the sort of reciprocitythat fair play requires, and it is interesting—though there is no spaceto trace the argument here—that these relevant aspects of justice arethe aspects of justice to which legality also has some important rela-tion (generality, for example).113

So once again, one can imagine an honorable rescue effort tryingto dispel the appearance of inconsistency as between the answers—(2a) and (2b)—that Hart gave to question (2). It is worth noting,however, that for most positivists, the separability thesis would still bein tension with even weak versions of (2a) when they are conjoinedwith (1a). For, as I understand it, the separability thesis is not onlysupposed to deny that whether or not a norm is law has conclusivemoral implications; it is also supposed to deny that it has any primafacie moral significance. The separability thesis is certainly not satis-fied by showing that although a norm’s being law has some moralimplications, those implications are not strong enough to settle thequestion of political obligation.

It is a pity that Hart did not himself explore any of this. Just inreviewing his work, we have uncovered an array of possibilities thatwould generate an interesting and nuanced set of relations—affirma-tive and negative—between the principles of legality and the positivistthesis of the separability of law and morality. Hart’s work is sugges-tive of all these possibilities, and it would have been very helpful hadhe or his followers seen fit to pursue them.

B. Reconciling Affirmative Answers

So far, we have dealt with various possible ways of reconcilingHart’s versions of (1a) and (1b) and various possible ways of recon-

112 See DWORKIN, supra note 91, at 190–216 (describing integrity as basis of politicalobligation); H.L.A. Hart, Are There Any Natural Rights?, in THEORIES OF RIGHTS 77,85–87 (Jeremy Waldron ed., 1984) (defending position that those who benefit from others’contributions have reciprocal obligation to contribute themselves).

113 See Waldron, supra note 107, at 184–85 (exploring relationship between justice andHart’s principle of fair play); see also Ronald Dworkin, Hart’s Postscript and the Characterof Political Philosophy, 24 OXFORD J. LEGAL STUD. 1, 23–26 (2004) (describing connectionbetween integrity and legality).



ciling Hart’s versions of (2a) and (2b). The problem for Hart,though—at least from the point of view of the separability thesis—isthe combination of (1a) and (2a). If one holds both that legality isamong the criteria for the existence of law and that legality hasinherent moral significance, it looks as though the criteria for the exis-tence of law include moral criteria. Yet this is exactly what positivistshave often denied.114

There are a couple of ways in which the embarrassment of thiscombination might be mitigated. Some positivists have acknowledgedthat the criteria for the existence of law or a legal system may be cri-teria that, as it happens, have moral significance. Joseph Raz observesthat the characteristic positivist claim “that what is law and what is notis purely a matter of social fact still leaves it an open question whetheror not those social facts by which we identify the law or determine itsexistence do or do not endow it with moral merit.”115 If (2a) wereheld to be true simply because the properties of a system of rule asso-ciated with observance of the principles of legality happened to beproperties that other moral principles made morally significant, then alegal positivist might be willing to live with that. We can call this con-tingent moral significance.116 If, on the other hand, the moral signifi-cance of the principles of legality were not coincidental, if it were suchthat the principles of legality themselves embodied certain moral prin-ciples or moral values, then the embarrassment to the positivist ofcombining (2a) with (1a) would be greater. We can call this noncon-tingent moral significance. On this account, legality would be a thickmoral predicate—like courage or honesty—and its criterial connec-tion with law would make law into a matter of partial moralassessment.

I am not sure how we should read Hart’s embrace of (2a). Some-times he seems to be saying, in consequentialist fashion, that it hap-pens to be a matter of fact that failures of legality, such asretroactivity, can harm human happiness. This seems to accord con-tingent moral significance to the requirement of prospectivity.117 Butat other times he seems to go much further than that—when, forinstance, he describes the requirement of prospectivity as “a very pre-

114 See, e.g., JOSEPH RAZ, PRACTICAL REASON AND NORMS 163–70 (3d ed. 1999)(arguing against natural law claims of necessary connection between law and morality).

115 JOSEPH RAZ, Legal Positivism and the Sources of Law, in THE AUTHORITY OF LAW,supra note 14, at 37, 38–39.

116 An analogy might be something like this: Secondary rules make possible the keepingof detailed archives; the keeping of detailed archives is a good thing, in the light of themoral value of historical knowledge; therefore, secondary rules have moral significance.

117 Hart, supra note 27, at 274.



cious principle of morality.”118 That sounds more like noncontingentmoral significance, and combining this version of (2a) with (1a) doespose some difficulty for the separability of law and morality.

Another way of looking at the relation between (1a) and (2a) isto ask whether the features or aspects of legality that are criteriallytied to law are also the very features or aspects of legality that havemoral significance. I do not mean to repeat the point I made at thebeginning of this Part—that there are many principles of legality, andsome may be tied criterially to law and others not, while some aremorally significant and others not. Even if the same principle oflegality is both tied to law and accorded moral significance, there isthe further question of whether the ways in which it is tied to lawmatch up with—or indeed literally are—the ways in which it is mor-ally significant. Consider the principle of generality, for example,which requires that all particular legal orders be based on generalrules or standards. One can imagine the following versions of (1a)and (2a) for generality:

(1a/ gen) Generality is criterially connected with law inasmuch aslaw cannot practicably operate without general rules. Our conceptof a legal system must be the concept of a system that can work, andwithout general rules a legal system cannot work.(2a/ gen) Generality tends to make a positive moral difference to asystem of rule inasmuch as it associates law with something likeKantian universalizability.

Juxtaposing these two claims may pose little embarrassment for theseparability thesis, because the pragmatic aspect of (1a/ gen) does notreally match up with the moral aspect of (2a/ gen). There are points inThe Concept of Law at which Hart comes close to embracing (1a/gen).119 In other places in that book, however, his account of gener-ality’s criterial link with law is not so distant from his account of itsmoral significance; this is particularly true in his discussion of the“peculiarly intimate connection” between the idea of law and the ideaof “treating like cases alike” and in his suggestion that the latter ideacontains “the germ at least of justice.”120 In this discussion, Hartseems to be clearly focused on exploring the possible moral signifi-

118 Hart, supra note 1, at 619.119 E.g., HART, supra note 12, at 21 (“[N]o society could support the number of officials

necessary to secure that every member of the society was officially and separately informedof every act which he was required to do. . . . Legal control is therefore primarily, thoughnot exclusively, control by directions which are . . . general.”); Hart, supra note 27, at 274(“[N]o society could efficiently provide the number of officials required to make them amain form of social control.”).

120 HART, supra note 12, at 157–67, 206.



cance of the very ways in which law presents itself in terms ofgenerality.

With the other principles of legality, the case for a connectionbetween (1a) and (2a) is even clearer. Many (if not all) legal posi-tivists regard it as definitional that the function of law is to guide con-duct.121 That law cannot, in general, guide human conduct unless itsdirectives are clear, public, prospective, practicable, and relativelyconstant relates directly (as Fuller noticed122) to the moral ideal ofrespecting the human capacity for responsible agency and self-monitoring. The idea that, if one is to rule human beings, one shouldwork with this capacity—rather than short-circuiting it throughmanipulation or terror—is an idea of considerable moral significance;a system of rule is better if humans are ruled in this rather than insome other way. Of course, a ruler may have reasons of his own fortrying to guide the conduct of his subjects (rather than galvanizing itin some other way), and those reasons need not themselves involvemoral respect for the dignity of human agency. No one is denying thatrulers may have nonmoral reasons for abiding by the principles oflegality. But that does not deprive the principles of their moral signifi-cance, nor does it mean that their criterial connection with law ispurely a result of rulers’ characteristic opportunism. Law itself maybe an enterprise unintelligible apart from the function of treatinghumans as dignified and responsible agents capable of self-control;unscrupulous rulers must make what they can of that fact when theydecide, for reasons of their own, to buy into the “legal” way of doingthings.

CONCLUSION

I have two conclusions. The first is to insist that Lon Fuller’s 1958response to H.L.A. Hart’s Holmes Lecture remains importantly sug-gestive for modern jurisprudence. Hart may have tried to create theimpression that Fuller’s response and his later book were hopelesslyconfused, but Hart himself—when he thought that no one waslooking—toyed with many of the positions that Fuller held. And the

121 See JULES L. COLEMAN, THE PRACTICE OF PRINCIPLE: IN DEFENCE OF A PRAGMA-

TIST APPROACH TO LEGAL THEORY 206 (2001) (referring to importance of law’s guidingfunction). For a discussion of the significance of this necessary feature of law with regardto the rule of law, see JOSEPH RAZ, supra note 14, at 210, 218.

122 FULLER, supra note 17, at 162:Every departure from the principles of the law’s inner morality is an affront toman’s dignity as a responsible agent. To judge his actions by unpublished orretrospective laws, or to order him to do an act that is impossible, is to conveyto him your indifference to his powers of self-determination.



record we have uncovered of his toying with these positions suggeststhat there is a lot more fruitful work to be done in this area.

The other conclusion is that jurisprudence—particularly positivistjurisprudence but, perhaps, the jurisprudence of its opponents too—isdisfigured and diminished by an obsession with the separability thesis,particularly when the separability thesis is stated in a very dogmaticand broad-brush form. The major achievements of Hart’s legal phi-losophy—particularly his attack on sovereignty123 and the commandtheory,124 and his own insight into the practice theory of rules,125 theinternal aspect of rules,126 and the distinctive structures of a legalsystem127—would be intact even if he had felt it necessary to abandonthe position that none of the aspects of social life that determinewhether a society has a legal system can have any inherent moral sig-nificance. In the hands of Jules Coleman and others, positivist juris-prudence has made great progress in exploring various possibilitiesthat are not ruled out by Hart’s own very particular formulation, inThe Concept of Law, of the separability thesis (“[I]t is in no sense anecessary truth that laws reproduce or satisfy certain demands ofmorality . . . .”).128 Coleman has shown that this formulation leavesopen various “soft positivist” possibilities and that this openness isquite fruitful for jurisprudence.129

The present Article has worked in a somewhat different direc-tion, exploring possible connections between law and morality vialegality (rather than via the contingent characteristics of particularrules of recognition or practices of interpretation). But the same gen-eral point may apply. The combination of what I have called position(1a)—that there is some criterial connection between legality andlaw—and position (2a)—that principles of legality do have moral sig-nificance—certainly challenges a very broad version of the separa-bility thesis. It does not, however, challenge the exact formulation

123 See, e.g., HART, supra note 12, at 50–78 (discussing failings of Austin’s doctrine ofsovereignty).

124 See, e.g., id. at 18–25 (rejecting command theory).125 See, e.g., id. at 254–59 (discussing and addressing criticism of practice theory).126 See, e.g., id. at 56–58 (distinguishing internal aspect of rules from “habit of

obedience”).127 See, e.g., id. at 79–91 (introducing primary and secondary rules); id. at 91–99 (pos-

iting primary rules of obligation and secondary rules of recognition, change, and adjudica-tion as characteristic elements of legal systems).

128 Id. at 185–86; see also, e.g., Jules L. Coleman, Negative and Positive Positivism, 11 J.LEGAL STUD. 139, 148–49 (1982) (“The separability thesis survives [if] not every conceiv-able legal system has in its rule of recognition . . . a clause [setting] out conditions oflegality for some moral principles, or if it has such a clause, there exists at least one con-ceivable legal system in which no principle satisfies [it].”).

129 See COLEMAN, supra note 121, at 67–68, 103–19, 151–52.



that Hart used in The Concept of Law—that is, the combination of(1a) and (2a) does not imply that laws necessarily “reproduce . . . [the]demands of morality,”130 though it does imply that some aspects ofwhat it takes to be a law do have moral significance. I think this is notjust a verbal difference but a genuine openness in the otherwise dog-matic commitments of legal positivism.131 And I hope that my fellowlegal philosophers will think it worth taking advantage of this open-ness to continue, perhaps more profitably, Lon Fuller’s considerationof the connections between the concept and the rule of law.

130 HART, supra note 12, at 186.131 See the fine discussion in John Gardner, Legal Positivism: 5 1/2 Myths, 46 AM. J.

JURIS. 199, 210 (2001). Gardner points out that positivists are not debarred from thinkingthat “unclarity, uncertainty, retroactivity, ungenerality, obscurity and so forth are demeritsof a legal norm”; he also says that legal positivists are “not debarred from agreeing withFuller that these values constitute law’s special inner morality, endowing law with its owndistinctive objectives and imperatives.” Id.

POSITIVISM AND LEGALITY: HART’S EQUIVOCAL RESPONSE TO FULLER · TO FULLER JEREMY WALDRON* Lon...

Documents

Transcript of POSITIVISM AND LEGALITY: HART’S EQUIVOCAL RESPONSE TO FULLER · TO FULLER JEREMY WALDRON* Lon...